ground detection and noise removal

370 views
Skip to first unread message

Seweryn Cieslik

unread,
Jul 6, 2016, 9:35:24 AM7/6/16
to LAStools - efficient tools for LiDAR processing

Hi all

 

I have had a look through the forum but I still have a few questions J

 

1. First of all how would you go about detecting  ground in areas of mixed terrain. I understand that presets are designed to deal with certain types of terrain but how would you go about an area let say on the boundaries of a big city (flat) with warehouses, supermakets and a hilly park or cliffs across the road. (images 1a\1b, 2a\2b)

 

I have been experimenting with different values and so far I cannot get a good compromise. So at the moment my workflow is as follows: detect ground with a large step (like metro), classify buildings (a bit hit and miss at the moment), then detecting ground again with the class 6 (buildings) locked. 

 

I wonder if anyone worked that way before. The issue is that when using, for example, the metro setting it finds the ground accurately around the big buildings but will fail to find the ground along parts of the slope on hills. Using a smaller step value will find the ground fine but will start classing buildings as ground (image 3).

 

 

2. Is there a way in lasnoise to efficiently remove spurious points from below the ground? The ideal way would be to create a TIN and then tell the tool to look for any spikes that are n stops down (or up) from the standard deviation. 

 

 

I hope my questions are making sense and I look forward hearing from you.   

 

I have attached some images for clarification.

 

 

Seweryn Cieslik

first pass_1a.png
first pass_1b.png
first pass_2a.png
first pass_2b.png
first pass_3.png

Martin Isenburg

unread,
Jul 23, 2016, 4:10:57 AM7/23/16
to LAStools - efficient command line tools for LIDAR processing
Hello,

if you have complex terrain with buildings next to steep mountains then I suggest to try the newer "lasground_new" in addition to "lasground" as this was particularly engineered for such scenarios.


Another strategy is to first classify the bigger buildings using lasground with a coarser step size and then run lasground again while ignoring the points classified already classified as buildings during a second lasground run with a finer step size:

lasground -i in.laz -metro -ultra_fine -o temp1.laz
lasheight -i temp1.laz  -o temp2.laz
lasclassify -i temp2.laz  -o temp3.laz
lasground -i temp3.laz -ignore_class 6 -town -extra_fine -o temp4.laz
lasheight -i temp4.laz  -o temp5.laz
lasclassify -i temp5.laz  -o out.laz

Spurious points below the ground are best removed with lasnoise.


And here is a small example workflow. I prefer to classify the points as noise (rather than removing them) so I can see *what* has been classified as noise and adjust the parameters if this was done too aggressively:

lasnoise -i in.laz -o temp1.laz
lasground_new -i temp1.laz -ignore_class 7 -city -ultra_fine -o temp2.laz
lasheight -i temp2.laz  -o temp3.laz
lasclassify -i temp3.laz -ignore_class 7 -o out.laz

For removing *low* noise aggressively have a look at this blog post that was written for point clouds from dense-matching phoogrammetry where clusters of low points seem to occur frequently in image areas that are strongly shadowed.


Regards,

Martin @rapidlasso

--
Download LAStools at
http://lastools.org
http://rapidlasso.com
Be social with LAStools at
http://facebook.com/LAStools
http://twitter.com/LAStools
http://linkedin.com/groups/LAStools-4408378
Manage your settings at
http://groups.google.com/group/lastools/subscribe

Reply all
Reply to author
Forward
0 new messages